Hi Greg,
On Wed, 2024-05-01 at 22:38 -0400, Greg Malysa wrote:
> Thanks for fixing this for me.
>
> > Confusion probably originates from Linux commit 685e444bbaa0
> > ("mmc: sdhci: Add ADMA2 64-bit addressing support for V4 mode"), but
> > the latter "V4 mode" was never ported to U-Boot.
>
> I have one of the rare platforms that does not support 64-bit ADMA2 V3
> so we also submitted 5359cd1135 ("mmc: Support 32-bit only ADMA on
> 64-bit platforms") to support that, but if we saw hardware that
> required v4 descriptors or someone insisted on it, we'd need to add
> another Kconfig for it. Do you think we should port the rest of the v4
> support? What about mirroring the kernel's dynamic behavior by
> checking the combination of capabilities and control registers to
> figure out which mode to use rather than hardcoding it?
current u-boot master contains ADMA_DESC_LEN value from 685e444bbaa0
without all the pre-requisites. My patch changes it to 685e444bbaa0^ state.
My patch will help to apply 685e444bbaa0^^^...685e444bbaa0,
but I'd prefer someone with some experience in this
area would do it, I'm personally not so sound in SD/MMC controller spec.
--
Alexander Sverdlin
Siemens AG
www.siemens.com
